comprehensive Psychological Assessment: A evel lop mental Psychopathology Approach for Clinical and Applied Research ALICE S. CARTER, SUSAN E. MARAKOVITZ, and SARA S. SPARROW MODEL OF COMPREHENSIVE pSYCHOLOGICAL ASSESSMENT 184 Multidimensional Components 184 TEST CONSTRUCTION AND SCALING ISSUES 197 Item Selection 197 Standardization 198 Reliability 198 Validity 199 Sensitivity-Specificity 199 METHODS 200 Reporting Source 200 Interviews, Rating Scales, and Questionnaires 200 Observational Methods 201 Projective Methods 201 In this chapter, we present a model of Comprehensive Psy- chological Assessment that has been developed to inform the assessment process in both clinical and research set- tings and has been influenced by the core tenets of develop- mental psychopathology. Comprehensive Psychological Assessment draws on the tenets of developmental psycho- pathology at each stage in the assessment process, from de- veloping and honing the question or questions to be addressed, identifying the components of the model that will be actively employed, choosing appropriate instrumen- tation, interpreting qualitative and quantitative informa- tion derived, and ultimately, when appropriate, making and implementing recommendations for prevention, interven- tion, and public policy. The Comprehensive Psychological Assessment model applies throughout the life span, but in this chapter we illustrate the components and application of the model for children and adolescents. Researchers and clinicians are accustomed to using hy- potheses and referral questions as catalysts for initiating the IMPLEMENTING THE MODEL: HYPOTHESIS-DRIVEN ASSESSMENT 202 Goals of the Assessment 202 Hypothesis-Driven Assessment Approach 202 Appropriate and Inappropriate Referral and Research Questions 203 Steps to Interpretation 203 RECENT ADVANCES AND DIRECTIONS FOR FUTURE RESEARCH 205 CONCLUSION 206 REFERENCES 206 assessment process. Following the refinement of hypotheses and questions, .the next step is to identify the relevant do- mains of functioning that need to be evaluated. Within the realm of clinical applications and research focused on devel- opmental psychopathology, it is extremely rare that only one domain of functioning (e.g., expressive language ability) is a sufficient target of evaluation. Because relative strengths and weaknesses in a single domain may or may not be iso- lated, investigators and clinicians generally recognize the importance of capturing patterns of functioning across mul- tiple developmental domains. For example, without knowl- edge of a child's general mental level or intellectual functioning, the meaning of below-average performance in one or more additional domains is ambiguous and potentially misleading. More specifically, a below-average score on a measure of receptive language that is administered as a proxy for cognitive functioning could capture low intellec- tual functioning. However, it could also represent a specific language delay or disorder, a visual attention problem, or a182 Comprehensive Psychological Assessment: A Developmental Psychopathology Approach for Clinical Applied Research broader pattern of developmental deviance such as that seen in Autism. Ultimately, patterns of strengths and weaknesses across domains have important implications for understand- ing developmental trajectories, diagnostic complexities, and contextual influences. The research or clinical question in combination with pertinent empirical literatures drives the selection of do- mains that will be the focus of any specific assessment ap- plication. Once relevant domains are identified, the careful selection of measures and choice of methods takes place. To capture the most coherent and accurate characterization of the intricacies of an individual's functioning, a currently ac- cepted standard is to utilize a multimethod, multi-informant approach (Johnston & Murray, 2003). In research applica- tions, shared variance attributable to different informants and methods can be modeled statistically using methods such as structural equation modeling (Kline, 1998), and shared variance that is attributable to nested contexts (e.g., family, classroom, school) can be modeled using hierarchi- cal linear modeling (Raudenbush, Bryk, Cheong, & Cong- don, 2000). At present, there are newly emerging but few empirical guidelines that clinicians can use for systemati- cally integrating information that spans not only multiple domains of functioning but also multiple informants and methods (Kraemer et al., 2003). Thus, in clinical settings, when informants present differing views about either the child's functioning or features of the child's ecology, the evaluator must predominantly rely on clinical judgment and intuition in the next phase of the assessment process: inter- pretation of obtained findings. In both research and clinical applications, interpretation of assessment findings can be aided by striving to form conceptuaIizations and impres- sions that are grounded in the tenets of developmental psychopathology. Developmental psychopathology is inherently interdisci- plinary in scope and encourages cross-fertilization of epis- temologies and methodologies. Cicchetti and Sroufe (2000, p. 256) elegantly state that what is central to the discipline of developmental psychopathology is the relevant ecologies that transact with the individual>, functioning over time. Moreover, we continue to advocate for understanding and recognizing the separate and inter- woven relations between current resources and adaptive functioning. Adopting this approach presents both chal- lenges and responsibilities, as clinicians and researchers must embrace multiple complexities simultaneously but still be able to generate meaningful conclusions. Consistent with one of the thrusts in developmental psychopathology, a major task for clinicians and re. searchers is to distinguish between adaptive and maladap. tive presentations. This can be done only with attention to developmental considerations and an understanding of broader contextual influences. While it is necessary to make use of knowledge about what is normative in each do- main at each age or stage, it is not sufficient. The selection of relevant domains and
View Full Document